Legal claims defining the scope of protection, as filed with the USPTO.
1. A method for clusterizing information, comprising a. determining similarity matrix for historical information and similarity matrix for current information; b. generating an aggregated similarity matrix (aggregated kernel); and c. applying evolutionary spectral clustering on the aggregated kernel to a content stream to produce one or more clusters.
2. The method of claim 1 , comprising changing the number of clusters.
3. The method of claim 2 , comprising scaling the similarity matrices.
4. The method of claim 1 , comprising linearly combining the similarity matrices to obtain the kernel.
5. The method of claim 1 , comprising determining a quality of current cluster result.
6. The method of claim 1 , comprising determining temporal smoothness.
7. The method of claim 1 , comprising generating evolutionary clusters.
8. The method of claim 1 , comprising defining a cost function to measure a quality of a clustering result on evolving information.
9. The method of claim 8 , wherein the cost function is defined using one or more graph-based measures.
11. The method of claim 10 , wherein CT represents a goodness-of-fit of the current clustering result with respect to historic data features.
12. The method of claim 10 , wherein CT measures a cluster quality.
13. The method of claim 10 , comprising determining a negated average association for evolutionary spectral clustering.
14. The method of claim 10 , comprising determining a normalized cut for evolutionary spectral clustering.
15. The method of claim 1 , comprising deriving corresponding optimal solutions.
16. The method of claim 15 , wherein the optimal solutions are relaxed.
17. The method of claim 1 , comprising clusterizing blog sites for community detection.
18. A method for clusterizing information, comprising a. determining a first similarity matrix from a historic cluster obtained from historic information; b. generating an aggregated similarity matrix (aggregated kernel); and c. applying evolutionary spectral clustering on the aggregated kernel to a content stream to produce one or more clusters.
19. The method of claim 18 , comprising changing the number of clusters and providing for insertion or removal of one or more nodes.
20. The method of claim 18 , comprising linearly combining the similarity matrices to obtain the aggregated kernel.
21. The method of claim 18 , comprising determining temporal smoothness.
22. The method of claim 18 , comprising generating evolutionary clusters.
24. The method of claim 18 , comprising clusterizing blog sites for community detection.
25. The method of claim 18 , comprising changing cluster numbers.
Unknown
November 9, 2010
Browse 5M+ US patents with plain-English claim translations and AI-generated analysis.